Gaining Ground, Second Edition: The Origin and Evolution of Tetrapods

GAINING GROUND

Life of the Past James O. Farlow, editor

This book is a publication of

Indiana University Press

601 North Morton Street

Bloomington, Indiana 47404-3797 USA

iupress.indiana.edu

Telephone orders 800-842-6796

Fax orders 812-855-7931

© 2012 by Jennifer A. Clack

All rights reserved

No part of this book may be reproduced or utilized in any form or by any means, electronic or mechanical, including photocopying and recording, or by any information storage and retrieval system, without permission in writing from the publisher. The Association of American University Presses’ Resolution on Permissions constitutes the only exception to this prohibition.

The paper used in this publication meets the minimum requirements of the American National Standard for Information Sciences—Permanence of Paper for Printed Library Materials, ANSI Z39.48-1992.

Manufactured in the

United States of America

Library of Congress

Cataloging-in-Publication Data

Clack, Jennifer A., 1937–

Gaining ground : the origin and evolution of tetrapods / Jennifer A. Clack. — 2nd ed.

p. cm. — (Life of the past)

Includes bibliographical references and index.

ISBN 978-0-253-35675-8 (cloth : alk. paper) — ISBN 978-0-253-00537-3 (e-pub) 1. Lungfishes, Fossil. 2. Amphibians, Fossil. 3. Leg—Evolution. 4. Paleontology-—Devonian. 5. Paleontology—Carboniferous. I. Title.

QE852.D5C57        2011

1 2 3 4 5    17 16 15 14 13 12

Contents

PREFACE TO THE SECOND EDITION

ACKNOWLEDGMENTS

LIST OF ABBREVIATIONS

1 Introduction: The Origin and Evolution of Tetrapods

2 Skulls and Skeletons in Transition

3 Relationships and Relatives: The Lobe-Fin Family

4 Setting the Scene: The Devonian World

5 The First Feet: Tetrapods of the Famennian

6 From Fins to Feet: Transformation and Transition

7 Emerging into the Carboniferous: The First Phase

8 East Kirkton and the Roots of the Modern Family Tree

9 The Late Carboniferous: Expanding Horizons

10 Gaining Ground: The Evolution of Terrestriality

REFERENCES

INDEX

Preface to the Second Edition

Since the first edition of Gaining Ground was completed, much has happened in the field of early tetrapod paleontology and in the wider world that has affected not only the ideas and conclusions presented in the first edition, but also what paleontologists are able to do with material, the techniques that they use, and how easily and quickly things can be done.

One of the most significant events worldwide has been the global spread of the Internet, which has not only allowed discoveries to be published more quickly, but has enabled rapid searches for material and references. The ease and speed of access to references by electronic means and Internet search have improved enormously, allowing information and citations to be readily found at the touch of a button. It’s hard to remember that in the late 1990s and the first couple of years of the 21st century, when the first edition was written, online connections were often slow and limited. Today, the ease and availability of electronic techniques has greatly affected many aspects of producing and delivering a wide range of scientific output. I personally have been most affected by my familiarity with software packages such as Photoshop, which permits manipulation of scanned images to produce diagrams. Some of the diagrams in the first edition were admittedly clumsy, and I hope that readers will find these improved in the second. Digital photography, of course, is another boon.

Technological advances have affected many areas of study, and paleontology is no exception here. One of the techniques that has improved in availability, cost, and degree of resolution is that of X-ray computed tomography (CT), or micro-CT scanning. This is becoming the technique of choice for examining new aspects of fossil material previously inaccessible. Software programs build the serial sections produced by scans into three-dimensional images that can be easily manipulated and dissected. These have become more sophisticated but at the same time more amenable to being run on a moderately powerful desktop computer, as well as becoming more intuitive to use. Such advances have allowed new questions to be framed and answered. One stage up from micro-CT scanning with X-rays is the use of a synchrotron. This allows minute examination of tissue structure inside fossil material, from which, for example, three-dimensional images of growth patterns of bone can be built. For really high-resolution scanning to get these results, at present, the limitation of this technique is the small size of sample that can be examined at any one time. Larger specimens can be scanned at lower resolution, often higher than with micro-CT machines, but results still depend on the geological makeup of the material.

Micro-CT images of fossils are increasingly being used in biomechanical studies using an engineering technique known as finite element analysis (FEA). This allows study of the relative degrees and directions of stress that a structure, such as a skull, can withstand, thus permitting function to be inferred. Although FEA has not yet been widely applied to very early tetrapods, such analysis of lower jaw function, for example, should help reveal what the implications are for the changes in dental patterns seen across the fish–tetrapod transition. Such work is in progress.

Digital imaging and suitable computer programs allow quantitative studies of, for example, skull proportions, or the way in which different bones contribute to skull structure in different groups. This study, known as geometric morphometrics, is being applied to groups of fossil tetrapods. It can help tease apart differences in shapes among a range of taxa to reveal phylogenetic or morphological relationships, and some of the results are included in this new edition.

New electronic techniques have also been brought to bear on climate modeling and atmospheric composition in past periods of Earth’s history, and these are highly relevant to understanding the late Paleozoic and events that occurred during that time.

Computers’ increased memory, faster processing power, and lower cost have allowed increasingly large data sets of taxa and characters to be used in phylogenetic analyses, and new search protocols and methods are beginning to provide alternative ways of processing the information. Some of these advances have been made in the field of molecular phylogenetics and in the incorporation of morphological and molecular data into combined analyses.

Cladistic matrices and the trees they generate are now being interrogated further in metadata analyses to produce hypotheses of overall evolutionary patterns: diversity and disparity curves, effects and results of evolutionary constraints or restrictions on taxa, episodes of character release (i.e., increasing the number of potentially varying characters) or decreased rates of character change, evolution of body size among clades, and many others (e.g., Ruta et al. 2006; Wagner et al. 2006; Laurin 2004).

The electronic age has seen a veritable explosion of information in the field of evolutionary developmental biology, or evo-devo. In its infancy in the early 1990s, it has more recently been asking questions in a phylogenetic context, incorporating ideas and data from phylogenies that have been drawn by means of both molecular and morphological data and that include fossil taxa. Paleontology has much to contribute to that endeavor, but similarly, researchers in evo-devo fields have been addressing questions on the basis of phylogenetically relevant species. Sharks, lungfish, and nonteleost fish are examples of these species, which have provided key information in studies such as the origin of limbs and digits.

Within the field of early tetrapod vertebrate paleontology, there have also been major developments of a more traditional kind. People have gone out and found more fossils. The number of recognizably different taxa of Devonian tetrapods has increased more or less exponentially since the mid-1990s, and discoveries are still going on. It is true that similarly explosive expansion has occurred in other fields of vertebrate paleontology as well, so my field appears to be part of a general trend. In fact, this probably applies to all aspects of scientific discovery, and a good part of the explanation is embodied in the electronic advances outlined above. But for the field covered in this book, there are particular areas in which major steps have been taken. The increasing information on elpistostegalian fish, like the discovery of Tiktaalik and micro-CT scanning of parts of Panderichthys and Ichthyostega, has added excitement, complexity, and some frustration to the field. The recent announcement of apparently tetrapod-like trackways in deposits 18 million years before tetrapod-like animals were supposed to exist is another surprise whose implications need to be assessed.

Two other fossil tetrapod groups have been the subjects of revived interest, generating new finds and new analyses. One of these is the temnospondyls, and in particular the group called dissorophoids. For some members of this group, we now have significant and useable growth series, and for others, we simply have far more relevant taxa. The lepospondyls are another group on which attention has been focused, from both morphological and phylogenetic points of view. This is one reason why readers will find much expanded sections on both of these groups and why the review of their relationships is much more complex and perhaps more confusing than it was. It reflects the efforts of key researchers in these various fields, which has been quite intense and difficult to summarize.

As I was writing the first edition, very little indeed was known about the environmental conditions in which early tetrapods lived. A review of the first edition commented on the lack of information about the faunas with which early tetrapods were associated. The reason for that was simple. Little comparative information was available at that time. Since then, several meetings and much exchange of information on the topic have allowed meaningful comparative faunal studies to be carried out for the first time. We are now a little closer to understanding the circumstances in which the fish–tetrapod transition took place. I have expanded these sections of the text accordingly. That said, our information is still limited to a few sites that are reasonably informative.

In short, early tetrapod paleontology continues to be a field of great excitement and possibilities. Some of the predictions of the first edition have been proved incorrect, some show early signs of success, and some have come about.

Acknowledgments

This work essentially represents a summary of my research career in early tetrapod paleontology. Throughout this period, I have worked with, been guided and helped by, learned from, and been supported morally and financially by a large number of people and institutions whom it is now my opportunity to thank.

I begin with my late father, Ernest Agnew, who unfortunately did not live to see my election as a Fellow of the Royal Society. He and my husband, Mr. Rob Clack, have both been instrumental in encouraging me on this road, and have helped reduce the rather long odds against my arriving where I am. My mother, Alice, died in 1983, and so was not able to see what she also helped me achieve. It is also a delight to acknowledge the encouragement and understanding of my late mother-in-law, Mrs. Molly Clarke.

I belong to the Panchen school of early tetrapod paleontology, and thanks must go largely to Dr. Alec Panchen for having taken me on as a mature research student after several years working in a provincial museum. My boss at that time, Mrs. Anna Meredith, deserves my eternal thanks for having encouraged me to take the plunge and return to academia, as well as providing the opportunity that led to that outcome. I returned to museums after my postgraduate years, although not in the way I expected. When appointed to a post in the University Museum of Zoology, Cambridge (UMZC), in 1981, in a sense I returned to my ancestral home, since Alec was a student of Rex Parrington, whose work in early tetrapod paleontology is the foundation of the collections and research effort of the UMZC.

During my subsequent career, several colleagues have been key sources of advice, discussion, intellectual challenge, and stimulation, as well as sources of specimens and general paleontological and social fun: Drs. Mike Coates, Andrew Milner, Angela Milner, and Tim Smithson, all Panchen school graduates; my own research students, but particularly Prof. Per Ahlberg, whose research overlaps in a most satisfactory and fruitful way. In the UMZC, I am indebted to Dr. Ken Joysey, a former director, for crucial help in mounting the first Greenland expedition, and to Prof. Michael Akam, the previous director and now head of the Zoology Department; and to the former Curator of Vertebrates, Dr Adrian Friday, and the former head of the Department of Zoology, Prof. Malcolm Burrows, for continuing support both for my research and my personal progress. Mr. Ray Symonds, the former collections manager of UMZC, has been a source of practical support; he remained good-humored in even the most trying of circumstances. Several people have assisted practically with the project over the years, including Ms. Rosie Rush, Dr. Nick Fraser, and Dr. Henning Blom, as a postdoctoral associate. Most recently, my collaborators on an early tetrapod locomotion project have been Dr. John Hutchinson (Royal Veterinary College, London) and Dr. Stephanie Pierce. Ms. Julia Molnar is our scientific illustrator, and she has helped me enormously with the figures for this book, especially the front cover.

My expeditions to Greenland could not have even begun had not Rob mithered me until I gave in, and nothing would have progressed without the help of Dr. Peter Friend of Cambridge’s Department of Earth Sciences, and the unwitting help of his former student Dr. John Nicholson. The expeditions were supported logistically by the GGU/GEUS (Greenland Geological Survey of Denmark) led by Dr. Neils Henricksen (Oscar) and facilitated by the late Dr. Svend Erick Bendix Almgreen of the Geological Museum Copenhagen (MGUH). I thank both of these for the part they played in our expeditions’ successes. Ms. Sally Neininger (now Dr. Thomas) and Dr. Becky Hitchin were two of my gallant field assistants in 1998. More recent collaboration with MGUH through the good offices of Dr. Minik Rosing, its former director, Dave Harper, Professor of Paleontology, and Dr. Gilles Cuny, Curator of Vertebrate Paleontology, have allowed my team to continue working on the Devonian tetrapods we find so rewarding and fascinating.

In the UK, much of my material has been provided by the collector Mr. Stan Wood, and I record here the debt British vertebrate paleontology owes to him.

Further abroad, many colleagues have helped me by generously discussing their own research and allowing access to collections, and providing support in more subtle ways. I particularly wish to thank Prof. Eric Lombard and Dr. John Bolt from Chicago, for ongoing collaboration and lots of gin and tonics, Profs. Bob Carroll and Robert Reisz in Canada, and Drs. Anne Warren and Susan Turner in Australia. Other museums and their staff have given me generous access to their collections and specimens: in the UK, Drs. Mike Taylor and Bobby Paton formerly of the National Museums of Scotland, Dr. Neil Clarke of the Hunterian Museum Glasgow, Dr. Steve McClean and Mrs. Sylvia Humphrey of the Hancock Museum, Tyne and Wear, Ms. Sandra Chapman of the Natural History Museum, London; in the United States, Dr. Dave Berman of the Carnegie Museum Pittsburgh, Dr. Ted Daeschler of the National Academy of Sciences in Philadelphia, Dr. Gene Gaffney of the American Museum of Natural History in New York, Dr. Farish Jenkins of the Museum of Comparative Zoology, Harvard; in Eastern Europe, Dr. Erwins Luksevics in the Natural History Museum in Riga and Dr. Oleg Lebedev of the Paleontological Institute in Moscow. Many other colleagues and friends helped directly or indirectly in my work on early tetrapods, and I thank them all.

The other side of my research interest, the evolution of hearing, was greatly influenced by two people, Dr. Arthur N. Popper, who encouraged me into the world of auditory neurophysiology, and Dr. Christopher Platt, my close friend and colleague. I thank them both for welcoming a paleontologist into their world.

My research has benefitted from funding from the following bodies, which I gratefully acknowledge: Natural Environment Research Council, UK; National Geographic Society; Newnham College; and the Department of Zoology, University of Cambridge; UMZC, Isaac Newton Trust Fund, Cambridge; Copenhagen Biosystematics Centre.

Specific thanks in the production of this book go to Dr. Chris Berry for help with Devonian plants, Prof. Mike Coates for help with Hox genes, Drs. Andrew Milner and Jason Anderson for helpful comments on Chapter 9, and my editors, Dr. Jim Farlow and Bob Sloan.

Finally, I would like to draw particular attention to the part played by my former preparator for more than a decade, Mrs. Sarah Finney, without whose patient, exquisite, and delicate work on the fossil material this work could hardly have happened. She also produced many of the photographs that enhance this book. Early tetrapod paleontology owes her a great debt of gratitude.

Abbreviations

Abbreviations of People and Institutions

GAINING GROUND

1.1. Family tree of the living tetrapod groups.

1 Introduction

The Origin and Evolution of Tetrapods

Approximately 380 million years ago, something strange and significant happened on Earth. That time is part of an interval of Earth’s history called the Devonian period by scientists such as geologists and paleontologists, but in more popular imagination, it is known as the Age of Fishes. The reason for this is that after about 200 million years of earlier evolution, the vertebrates—animals with backbones—had produced an explosion of fishlike animals that lived in the nearshore lagoons, river estuaries, and lakes of the time. The strange thing that happened from the middle to the later parts of the Devonian period is that some of these fishlike animals evolved limbs with digits—fingers and toes. Over the ensuing 350 million years, these tetrapods gradually evolved from their aquatic ancestry into walking terrestrial vertebrates. These have dominated the land ever since this initial explosive radiation allowed them to colonize and exploit the land and its opportunities. The tetrapods, with their limbs and fingers and toes, include ourselves as humans, so that this distant Devonian event is profoundly significant for humans as well as for the planet.

Today, the modern descendants of these early pioneers are divided into two major groups (Fig. 1.1). The modern amphibians include frogs (anurans—jumping, tailless amphibians), salamanders (urodeles—tailed amphibians), and caecilians (apodans—elongate, limbless amphibians). The modern amniotes include mammals (usually animals with fur and that produce milk, humans among them), turtles, crocodiles and their relatives the birds, and lizards and their relatives the snakes. (Note that the term reptile can be used to include all amniotes except mammals, provided birds are included in this group.) In practice, tetrapods include any animal with four legs or whose ancestor had four legs. (The situation is a bit more difficult with fossils, as will be shown.) These two major radiations of vertebrates also had many relatives that are now extinct, notably the dinosaurs, whose closest living relatives are the birds.

Because becoming terrestrial was apparently a slow process, I have called this book Gaining Ground, to suggest that it was achieved with some difficulty. At the same time, this evolutionary development has led to enormous innovations in evolutionary terms, and I hope to suggest not only something of the breadth of possibilities that it opened up, but also the contingent nature of the transition. In other words, much of this change occurred because of happenstance—being at the right place at the right time. It was not a directed process.

The origin, early evolution, and relationships of tetrapods form the focus for the interaction of several disciplines. Paleontology (the study of fossils) and the related studies of paleoecology, taphonomy (how the creatures died and became fossilized), and paleobiogeography (where the creatures lived and how they were distributed in time and space), as well as modern zoology, anatomy, physiology, and developmental and molecular genetics, all contribute various aspects to the understanding of past life. Most people are aware that at some stage creatures crawled out of the water and came onto land, and so can relate to the contents of this book. I hope this book will show how much more can be said than that—and how much more there is to know.

The study of the origin of tetrapods has gone through many phases in its history, but none has been more exciting than that of the present day. Over the last few years, more fossil material from this crucial period has been unearthed than at any time in the past (this is even more the case than when the first edition of this book was being prepared 10 years ago), and these discoveries have helped to reshape ideas about when, where, how, and even possibly why the transition occurred at all. In the not-too-distant past, there was almost no fossil material, and ideas were based largely on informed guesswork. Speculation was intense, and as is often the case, this speculation was in inverse proportion to the amount of data. To be truthful, there is still not a large amount of real data, so speculation is still active, and whatever is concluded today may be overturned by discovery of a new fossil tomorrow. In some ways, this is to be hoped for, because only in that way can guesses be falsified and tested as scientific hypotheses. Indeed, much of this has gone on since the first edition of this book, as subsequent chapters will testify.

This book tells the story of the evolution of tetrapods from their fish ancestry and puts the sequence of events into its ecological context. The story is founded on an understanding of the evolutionary relationships between tetrapods and their fishy relatives—their phylogeny—and traces the family tree of tetrapods from its roots to the point at which the major groups of modern tetrapods branch off from its original trunk. The tetrapod family tree is in fact more like a bush, with several main branches, some of which have died out during the course of evolution and some of which have become large and important from small beginnings.

This book looks at the changes that occurred in the transition from creatures with fins and scales to those with limbs and digits in an attempt to understand how and when the changes occurred, and to do this, it is necessary to understand something of the anatomy of the animals involved. Chapters 2 and 3 are devoted to these parts of the story. Chapters 4, 5, and 6 set out what is currently known of the earliest tetrapods and their lifestyles. By careful analysis of what is known of them from fossils, and by comparison with modern animals that live at the transition between water and land, it may be possible to understand a little of how the early tetrapods worked as animals. After the tetrapods had become established, they radiated into a range of forms requiring modifications to the original tetrapod pattern. Chapters 7, 8, and 9 carry the story forward from the origin of tetrapods to their ultimate conquest of terrestrial living. The final chapter draws together some of the threads that have been taken up in the preceding chapters and shows how they impact the study and understanding of tetrapods today.

A cautionary note should be added here. Many of the skull reconstructions, derived from the literature, have freely duplicated left or right sides to present a complete, but artificially symmetrical, picture. These reconstructions are therefore not suitable for use in morphometric studies. Also, in many cases, the dermal ornament on the skull roofs has been omitted.

I hope that this book brings the excitement of this field of study to a wider public, shows something of how paleontology progresses and what it can and cannot do, and, of course, most importantly, shows people a little more of how they fit into the broader picture of evolution.

The Geological Framework

Tools of the Trade: The Geological Framework, Fossilization, and Family Trees

To put the evolution of terrestrial tetrapods in its context, it is necessary to have an understanding of Earth’s history in general outline. This is not the place to discuss dating methods or techniques of stratigraphic correlation. These can be found in readily available geological textbooks such as that by Briggs and Crowther (2001) or Raup and Stanley (1979). However, it is necessary to explain the approximate dates and approximate lengths of time over which the story takes place, so that it can be put in the context of other major events in the story of evolution.

The geological column is the name that scientists give to the succession of times, dates, and names into which Earth’s history is divided. There are several ways of expressing this. It can be expressed in a way that accords each time interval a space proportional to its length, usually as a vertical column, always with the oldest at the bottom. Or it might be depicted as a sort of clock face. The problem with this method is that only a small proportion of Earth’s known history is represented by an abundant fossil record. The planet is estimated to be about 4,500 million years old, and the first signs of life (fossil bacteria) are dated at about 3,500 million years. Complex multicellular animals first appear commonly in the fossil record only about 550 million years ago, so that to use the clock face method has practical problems, in that most of it would effectively be empty. Another way is simply to set out the list of dates and names in their relative order, again with the oldest at the bottom, as shown in Figure 1.2. The numbers show the dates of the boundaries between the divisions and the lengths of time they have lasted.

The idea that the Earth is as old as this is a relatively recent one, dating back only to the early 19th century, and its appreciation has changed the perspective from which we view our place in its history. The concept has been called deep time. As an example, one of the important factors that study of deep time reveals is the complexity of climate change through Earth’s history, culminating in the appreciation of the possibility of human-induced global warming. This would not be possible without study of the Earth’s climate over the past few million years. Talking of perspectives, when considering the period in Earth’s history covered by this book, climate changes far more radical than recent ones are obvious. If global warming continues as predicted, eventually the climate may become something like it was about 15 million years ago in the Miocene period, but it will still be a long way from that which current study suggests prevailed when early tetrapods were alive.

1.2. Timescale showing the time of origin of major groups, and other events in Earth’s history. The shaded area indicates the period covered in this book. Dates from Gradstein et al. (2004).

The interval for which there are abundant fossils in the rocks is called the Phanerozoic, meaning visible life, and it represents a time of about 600 million years. The Phanerozoic is divided into three eras, originally named according to what proportion of its biota resembled that of the modern world. These divisions are named the Paleozoic (ancient life), Mesozoic (middle life), and Cenozoic (recent life). The ages are divided into periods, and the periods into stages. To a large extent the boundaries of the divisions are based on the fossils of animals and plants that lived at that time, although the names they receive do not necessarily reflect this. Names of the stages, for example, are often based on places where the representative strata were first found or where they are most clearly seen. These large time periods, rock sequences and their names, and the basic faunal complement of each were worked out during the 19th century, and they have not changed very much since then. The subsequent decades have seen a process of refinement, increasing resolution of time intervals, and precision of dating and correlation between sequences in different parts of the world.

Geologists and paleontologists usually date their finds by reference to an independently produced and well-established timescale. One of the most recent in use at the present time is that by Gradstein et al., published in 2004, and has used many different and complementary techniques in its construction. It is enormously detailed for the whole Phanerozoic.

The story of the origin and early evolution of tetrapods and the time frame of this book takes place almost entirely within the later part of the Paleozoic, during the Devonian, Carboniferous, and Permian periods (Fig. 1.2). More details of the stages into which these periods are divided are given in the relevant chapters. The story as we have come to understand it really begins about 380 million years ago, although some recent work suggests that an earlier date is likely. Some chapters set the scene by describing the history of plants and animals that were already present as the tetrapods started their evolutionary journey. It takes the story through about 132 million years to a time about 248 million years ago as the Permian period comes to a close. For comparison, the first dinosaur is dated at around 225 million years, while the last died out 65 million years ago, a comparable period of time. The earliest tetrapods that feature in this story are nearly twice as old as the oldest dinosaur. Humans can trace their lineage back to a split from the common ancestor of apes and humans about 5 million years ago, while Homo sapiens as a species is currently reckoned to be about 100,000 years old.

Fossils, Fossilization, and Techniques of Study

The only means of finding out about animals and plants that lived so long ago is from fossils. These are the preserved remains or traces of these ancient organisms, and to understand the story more fully, it is necessary to look at how and under what circumstances these remains are preserved and discovered.

Types of fossils can be categorized in a variety of ways according to what is preserved, or how it is preserved. What is preserved is usually the harder parts of an animal or plant, so for example bones and shells make good fossils. This type of fossil is often called a body fossil, to distinguish it from another category, that of trace fossil. Trace fossils are preserved impressions of features that an animal (usually an animal, although occasionally plants leave trace fossils too) has made—for example, footprints or burrows. Body fossils show the anatomy of the plant or animal it preserves, and from this it is possible to work out something of its evolutionary relationships and functional morphology. Trace fossils can sometimes be even more telling in that they can provide information about the behavior of the animal and clues to its lifestyle that body fossils cannot give. Both kinds of fossil are known in the story of the origin of tetrapods.

Body fossils are preserved in many different ways. Usually the process involves water to a large extent, and fossils of aquatic animals are much more common than those of terrestrial animals. Fossils of terrestrial animals are usually found only when the creatures’ bodies have been accidentally washed into bodies of water. Generally, the animal sinks to the bottom of the lake or sea and the soft, fleshy parts usually decay quite rapidly. The remains may be scavenged by other animals and disintegrate so that bones become isolated, but gradually they are covered with sediment that over the millennia hardens to preserve the bones. Most favorable to preservation are deep, still waters where the sediment particle size is small. Sediments can then take up small details of the bones or shells by filling in even the smallest crevices, and then not being disturbed again. If decay happens more slowly—for example in water low in oxygen, where predators are few and bacterial action slow—the carcasses may be preserved in a more complete form.

The remains may be subsequently preserved in a variety of ways. They may be more or less unaltered. Shells made of calcium carbonate, for example, may retain the same chemical structure they had in life if they are preserved in limestone. Bones, which are the main concern here, are formed of calcium phosphate, which may also be unchanged chemically in many instances. However, bones are not solid but have pores or spaces in them, for blood vessels, nerves and fluid or even air, to allow them to grow and to make them lighter and stronger than solid bone would be. During fossilization, these spaces are often suffused with solutions of chemicals that later precipitate out and harden, so that the fine internal structure of the bone is preserved. This is why fossil bone is often much heavier than recently dead bone. Figure 1.3 shows sections through part of the skull of Acanthostega, a Devonian tetrapod, to show how well the internal structure of the bone can sometimes be preserved. This sort of detail can be used to explore aspects such as cell size, growth rates, annual or seasonal cycles, and microarchitecture.

Sometimes even the hard parts of the bone are dissolved away, leaving a natural mold of the original. This may be preserved as a space in the rock, or it may be replaced by another mineral to form a natural cast. Other hard parts of organisms that can be preserved are the teeth of vertebrates, whose enamel layer is often completely unaltered chemically even in ancient specimens, the chitin of arthropods, the silica spicules of sponges, or the woody (cellulose) parts of plants.

1.3. Photograph of sections through a skull of Acanthostega (MGUH f.n. 1604), shown on centimeter graph paper for scale. The palate and both lower jaws can be seen in the sections, but the skull roof is missing. The enlargement below shows dentary and coronoid with their teeth and the denticles on the prearticular. Photograph from UMZC archives.

Occasionally the remains have been buried and preserved so rapidly or in such unusual circumstances that decay has not had time to occur fully. Such occasions may result from a flash flood, inundating a river basin, or by an underwater avalanche in which soft sediment pours rapidly down an incline onto the organisms below. In some exceptional cases, even the softer parts of the organism may be fossilized. Some fish fossils, for example, retain details of the gill structure, and in such cases, muscle tissue shows such fine detail that individual cells can be identified. More commonly, soft tissue may be colonized by bacteria in the decay process (e.g., Fig. 9.15), and they may take up the shape of body outlines, fur or feathers, or gut contents. In recent years some spectacular examples of soft tissue preservation have come from China. Late Jurassic and Early Cretaceous dinosaurs with feather impressions have been found in large numbers, alongside many species of true bird, documenting the relationships between these groups of animals (Xu et al. 2003; Hou et al. 1996). From the same geological formations, early mammals with fur (Luo et al. 2007), salamanders, fish, and insects showing remains of color patterns have also been found (Chang et al. 2003). Even Early Cambrian and Precambrian fossils, also from China and preserved in exceptional circumstances, have provided new evidence not only of some of the earliest vertebrates (Shu et al. 1999, 2003), but also of embryonic animals consisting only of minute balls of cells (Donoghue et al. 2006).

For very ancient fossils such as those from the Paleozoic, the risk is high that during the period from their preservation to their discovery the fossils will have been altered in many different ways: by dissolution by acids that remove calcium and allow bones to bend into unnatural shapes without breaking; or by rock movements, faulting, or heat from igneous activity, which can distort and fracture them; or by exposure of the sediments to erosion in the air, which may remove or degrade them. Figure 1.4 shows a skull of Acanthostega distorted by crushing and twisting (compare this with the reconstruction based on examination of several specimens shown in Chapter 5, especially Fig. 5.21). Fossils may have disappeared when their sediments were subducted under other continental plates by tectonic movements. In short, fossils from earlier periods in Earth’s history are usually rarer and less well preserved than those of more recent periods.

Sometimes fossils can be preserved without the aid of waterborne chemicals or sediments. Amber is a resin produced by certain trees that occasionally traps insects or even small vertebrates such as frogs or lizards. When the amber is eventually hardened and preserved (again, usually in a water-laid sediment), access is even sometimes possible not only to the molecular structure of the animals, but also the chemical composition of the atmosphere in which they lived. Unfortunately for the story of early tetrapods, amber-producing trees had not evolved in the Paleozoic.

Sometimes animals die in very arid conditions, such as deserts, and their bodies become mummified (completely air-dried) before being buried in sand. When the sand hardens to sandstone, the soft tissue is sometimes preserved as well as the bones. Several kinds of dinosaur skin are known from this type of preservation, but so far, no early tetrapods have been found with soft tissue preserved in this way. Further details of how animal remains can be fossilized can be found in general textbooks.

When a fossil is found, it often remains in its surrounding rock, called matrix, and this has to be removed before study of the animal can begin. This special skill, called preparation, requires dedication and patience because it may take many years to complete a large or difficult specimen. Preparation may be done by mechanical means, such as a fine needle used under a binocular microscope, or by chemical means, such as dissolving away matrix by use of dilute acid. Sarah Finney (Fig. 1.5) used mechanical tools to remove the matrix from the fossils from Greenland. In this case, a dental mallet was used to chip tiny particles of rock off the fossil. The handpiece converts the drill action of this dentist’s apparatus into a reciprocating one, and both the throw and rate of the action can be adjusted. Some matrix that adheres strongly to the fossil may require finer work with a handheld mounted needle (or pin vice), as she is holding in the photograph, to remove the last particles grain by grain. For large tracts of bulk matrix, a pneumatic pen, like a tiny road drill, is the tool of choice. A skull of Acanthostega was bisected by the diamond wire saw to allow preparation of the internal surface, exposing the palate and internal surface of the lower jaw (Fig 1.6). The wire in this case had a diameter of only 0.3 mm, so very little of the original specimen was lost in the process.

1.4. Two photographs of Grace (MGUH f.n. 1300, Acanthostega). Anterior is to the right. Top, right lateral view; bottom, dorsal view. Scale bar = 10 millimeters. Photographs by S.M.F.

In the case of natural molds, specimens may be prepared by using a form of silicone or latex rubber to make casts from the original. Sometimes bone is badly preserved, in which case it is occasionally more useful to dissolve away the bone artificially and use a rubber peel from the resulting mold to perform the study. Some of the Carboniferous fossils described in Chapter 9, such as those from Linton, Ohio, are treated in this way. The surface detail obtained by this method is sometimes so fine that images produced by a scanning electron microscope can give valuable information.

1.5. Sarah Finney, the preparator who has done most of the work on the Devonian tetrapod Acanthostega and many of the early Carboniferous forms described in this book. The photograph is taken in the Paleontological Laboratory in the Department of Zoology, University of Cambridge, UK. Sarah is holding a skull specimen of Ichthyostega, and in the background is a life reconstruction of Acanthostega by Mike Coates. Photograph from UMZC archives.

Good, articulated (i.e., those still joined together) fossil skeletons of Devonian and Carboniferous tetrapods, some of which are described in this book, are rare. Only a handful of animals are represented by more or less complete skeletons; the rest are represented by individual bones such as lower jaws, partial skull roofs, or limb fragments. Usually the matrix in which they are preserved is hard, and the specimens have taken many hours of effort to extract. Occasionally, this general rule is broken: some early tetrapod fossils from the Baltic states are preserved in sandy material so soft that it can be washed off. The bones from these localities are three-dimensional and pale in color, so that they look almost like modern bone. The downside of this is that because the consolidation of the matrix is so poor, so is the consolidation of the fossil, and this creates a problem of conservation. It is difficult to extract them for the reverse reasons, in that they frequently fall apart without great care and use of consolidating plastics.

1.6. Photograph of Grace (MGUH f.n. 1300, Acanthostega, same specimen as in 1.4) showing the palatal surface prepared. Note the details of the teeth on the palate and the gill bars. Anterior is to the left. Scale bar = 10 millimeters. Photograph by S.M.F.

A newly developed technique, increasingly used since the late 1990s, is that of computer-assisted tomography, or micro-CT, in which a specimen is scanned in three dimensions by X-ray or other ionizing radiations. This technique, originally developed for medical purposes, has been refined for use in materials science and other physical studies, such that fine details of rock structure can be detected. The medically applied machines cannot usually use radiation of high enough energy to penetrate rock, although they have been used with great success in some cases, especially where the rocks and the enclosed bones are of contrasting densities. The rock-penetrating machines can now in many cases even cope with material containing iron compounds. Nonetheless, the material of Ichthyostega (Fig. 1.7), as illustrated in Chapters 5 and 6, was described as a challenge by the expert team at the University of Texas CT scanning unit. CT scanning can also be used for obtaining extremely fine detail from tiny specimens, allowing views of the internal structures of otherwise inaccessible parts, even distinguishing different cells and tissues at a histological scale.

An advance in the field of tomography has been the development of very fine-scale scanning with a synchrotron. Millimeter-scale portions can be scanned at quite large resolution, although one problem in dealing with larger specimens is that it is sometimes difficult to know where exactly the scan has been taken. The synchrotron can see almost to the cellular level, so that it is possible to build three-dimensional pictures of the internal microarchitectural structures of bone. In this way, successive growth surfaces have been revealed in bones showing changes to their shape and construction. Even the directions and areas of muscle fiber attachments can be made out, allowing muscle volume and direction of operation to be inferred. Larger specimens can also be scanned, although at lower resolutions, giving somewhat less detail. In some cases, however, depending on the characteristics of the matrix, better detail can be obtained using the synchrotron than with a more conventional CT machine.

1.7. Micro-CT scans of Ichthyostega. (A) Specimen of right half of a skull that has been scanned; lines show location of sections in (B) and (C). (B1) and (B2), respectively, photograph and interpretive drawing of more anterior section. (C1) and (C2), respectively, photograph and interpretive drawing of more posterior section. (D) Three-dimensional model of rear part of skull. (E) Interpretive drawing of model. Scale bar is for (A) only.

The advantages of CT scanning techniques are that they do not require full mechanical preparation of the specimens or they can act as a guide to how to prepare such material. The disadvantages of the technique are that first, it is still relatively expensive, and second, further expense is sometimes required to obtain the sophisticated software necessary to reconstruct the CT images into three-dimensional models, and to pay for the time of the researchers who are able to perform this process. It takes a lot of skill and knowledge to translate what may be a poorly resolved image into a credible model. Even with these new techniques, a level of interpretation is still involved in understanding the material.

Once a three-dimensional image has been obtained in this way, under certain conditions, it may be possible to investigate it more deeply. For example, a 3-D micro-CT scan reconstruction of a skull can be subjected to virtual analysis by engineering programs that allow the researcher to assess the stress forces that the skull can withstand. Finite element analysis has been used in this way to investigate dinosaur skull mechanics, although as yet not much has been done on early tetrapods. That is in part because the material is not always very amenable to scanning, or that it is too crushed to make a good 3-D model from.

Such new techniques have allowed completely new questions to be framed and answered, taking the investigation of fossil material to a more fundamental level.

Once the fossil is obtained and has been prepared and analyzed as far as it can, the time has come to describe it. To pass on the discovery to other researchers, so that they can use the information, is one of the most important parts of the scientific process. With fossils, other workers must have access to an accurate description of the material because many fossils are unique and are located in museums in distant parts of the world. The description is usually presented in a professional journal, in which a good clear verbal description is given alongside illustrations of the actual specimens and interpretive drawings and diagrams that explain how the describer sees and understands them. Ideas about the significance of the material, how it fits in to current knowledge, and what new hypotheses it challenges or presents are an important part of the scientific communication that results from such work.

Understanding Phylogeny

One of the first clues to the idea that life on Earth had evolved by a gradual process of change over time was the discovery that animals and plants could be placed in groups showing a hierarchical order. Small groups of forms showing detailed similarities could be placed with other small groups into larger ones showing more generalized similarities. The arrangement resembles a family tree, and it was this that first suggested the idea of descent with modification—that is, the realization that the similarities existing between more and more inclusive groups of animals result from ancestor–descendant relationships. Although the hierarchical classification of animals and plants had been recognized since the 18th century, and the idea that animals had changed throughout time had been suggested several times, what this meant remained mysterious and debatable until the mid-19th century. At that time, Darwin and Wallace finally realized that it could be explained by genealogical relationships, and they put forward compelling evidence for the idea. They also independently suggested a mechanism—known as natural selection—that could have brought about the changes. Everything that has been discovered since in modern genetics and developmental biology has only served to strengthen the evidence for this kind of relatedness among animals and plants. Many recent books discuss the patterns and processes involved in descent with modification, such as that by Jones (1999). This book updates Darwin’s original observations and expands the evidence he presented to include such topics as genetics, development, and plate tectonics. Others, perhaps more technical, are those by Barton et al. (2007), Futuyama (2005), and Ridley (2004).

The groups are given names to express these relationships. A species is the smallest group, usually meaning a group of organisms that can and do reproduce with one another. (This is not possible for extinct animals, of course, so paleontologists try to use other criteria. It is sometimes very difficult, especially if only fragments of an animal or plant are known.) Each species is given its own unique name. Species are grouped into genera (singular, genus) and share a generic name. Genera can be grouped into families, families into orders, and orders into classes (Fig. 1.8). In practice, in recent years, categories higher than genus are less used than formerly because they are in many ways incompatible with new methods of classification, such as cladistics.

An animal or plant species will have two names—a binomial. The first is the generic name and the second is the specific name. For example, one of the animals that features in this book is Panderichthys rhombolepis. Panderichthys is its generic name and rhombolepis is its specific name. Other species of the genus Panderichthys exist, such as stolbovi and bystrowi. Usually the first scientist to describe a new species will name it, unless it turns out to be a member of a genus that is already known, in which case the discoverer will only have to think up a new specific name. Each combination is unique to a species, and the name is designed to reflect something about the animal, such as who found it, where it came from, or some interesting feature of its anatomy. Scientists who name animals often have great fun doing so, although the international rules preclude facetious or vulgar names. Recently there have been suggestions to abandon this system because it was not designed to fit with recently developed methods of working out relationships between groups. However such suggestions are not universally accepted, and it remains to be seen whether they will stand the test of time and usefulness (e.g., Benton 2000; de Queiroz and Gauthier 1992; and other references listed in Chapter 3). Defining and naming higher categories is not always straightforward. Whether groups should be defined by reference simply to other taxa of apparently stable taxonomic relationships, by reference to living taxa, or by possession of certain suites of characters remains controversial. Examples of this are given in Chapter 3.

To understand the course of evolution, the evolutionary relationships between different groups must be worked out. Phylogeny is the name given to the evolutionary relationships and history of the animal in question, and the practice of working this out is called phylogenetics. The system of groupings in which an animal is placed is called its classification, and one of the main goals of evolutionary studies is to make classification reflect phylogeny.

Two major problems are encountered here. The first is that of discovering the best way of working out the phylogeny, and the second is that the phylogeny that is finally decided to be the best by scientists is often at odds with the classification used in everyday language.

1.8. Chart showing Linnaen hierarchy of species grouped into genera, and genera into families.

In recent times, one method for working out phylogeny that has become widely accepted is called cladistics. Named from the Greek word for branch (a clade), cladistics has some fairly strict rules about how to judge relationships between organisms, and it has its own language in which to express them. Often, commonly used words have a subtly different meaning in cladistics from those in everyday use, which may be confusing for the lay reader. One of the main rules is that only features (characters) that the groups (taxa) share uniquely (shared, derived characters) should be used to assess the relationship between them. This is in contrast to using characters that are more generally present in the larger grouping to which the organism might belong (shared primitive characters).

To give an example of this, in a classification of dogs and cats, possession of fur would not be a character used to unite them into a clade, because fur is found in many kinds of other animals. Fur is a character unique to a much larger group, the mammals. Dogs and cats instead share features of dentition only seen in members of the mammalian order Carnivora. Recently, it has emerged that whales show some specialized characters of the inner ear and anklebones that are otherwise found only in ungulates—cows, sheep, hippopotamuses, and their relatives. It is thought the most likely explanation is that they arose in a common ancestor of all these groups and have been inherited by all the descendants (Fig. 1.9). Another example might be to consider the three modern amphibian orders: frogs, salamanders, and caecilians. The fact that frogs and salamanders have legs, while caecilians do not, would not be used to suggest that frogs and salamanders were more closely related to each other than either one is to caecilians (Fig. 1.9). Legs are the common property of tetrapods (the main subject of this book) and have been lost by the ancestors of caecilians. (Indeed, some fossil caecilians are now known that have legs; Jenkins and Walsh 1993). Modern amphibians share a number of soft tissue characters, such as mucus glands in the skin, that are not found in other tetrapods.

1.9. Chart showing use of derived characters in phylogeny: even though frogs and salamanders all have legs, it is not these characters that are used to unite them, but a suite of others such as specialized receptors in the inner ear and glands in the skin unique to all modern amphibians; although whales lack legs, they are more closely related to ungulates such as sheep and cows than to carnivores, as shown by shared characters of the ear that are only found in that clade.

Problems arise with this method for two reasons. One is the means by which primitive characters are distinguished from derived ones, and the other is how genuinely shared derived characters (homologies) are distinguished from similarities derived independently (analogies). An example of an analogous character might be the lack of limbs in snakes and caecilians. Such characters may show something about lifestyle, but nothing about relationships.

These problems can get very complicated, especially when large numbers of extinct taxa are involved. In such animals, there are usually many characters for which no data exist, and many characters show conflicts (incongruence) in their distribution. In other words, the distribution of primitive and derived characters does not always overlap cleanly and simply in the way one might expect. This is partly because many of the taxa in question have had separate evolutionary histories, often of very long duration, during which each has undergone its own modifications. The answer these days is to use a computer. A table is drawn with the taxa listed along one axis, usually the columns, and the characters along the other, usually the rows. Each resulting square is filled in, showing the state of each character found in each taxon. The resulting table is called a matrix, not to be confused with the rock matrix enclosing fossils. Then the computer can find the branching arrangement (tree) of the taxa that involves the least incongruence and implies the fewest evolutionary changes (the most parsimonious). The tree or trees that are discovered represent the best available hypotheses of relationship drawn from the data used, but they are very much provisional. They will be subjected to much testing and are quite likely to be overturned by the addition of new data. Developments in cladistic analysis have recently included new methods that use likelihood or probability rather than parsimony as a basis for building phylogenetic trees. For morphological data, the results of this kind of analysis, known as Bayesian analysis, are not usually too far different from those obtained by parsimony analysis. In cases where an analysis includes extant forms, phylogenies based on molecular sequences can be used for comparison, although for early tetrapods, results from such studies have been ambiguous (see Chapter 3).

When a tree like this is drawn up, the task is then to name some of the branches. Because the animals are believed to be united into a clade by shared derived characters, all the members of the clade should be united by the same name to express this relationship, rather than arbitrarily split off some members from their closest relatives. This has led to some problems. For example, most paleontologists now believe that birds are most closely related to a group of small theropod dinosaurs (Fig. 1.10). In other words, birds are dinosaurs, evolutionarily speaking. Therefore they should not be promoted to a higher taxonomic status than dinosaurs and distinguished by being called a new class of vertebrates. In old classifications, this was what happened. The new method takes the evolutionary relationships of an animal group to be more important in its classification than its physical appearance.

A similar problem directly concerns the animals in this book. Because I will be dealing with a group of animals (tetrapods) that evolved from within a group of fishes (the lobe-finned fishes or sarcopterygians), technically, tetrapods are members of the